Skip to main content

Full text of "The Latin Concordance of Dante and the Genuineness of Certain of His Latin Works"

See other formats

Early Journal Content on JSTOR, Free to Anyone in the World 

This article is one of nearly 500,000 scholarly works digitized and made freely available to everyone in 
the world byJSTOR. 

Known as the Early Journal Content, this set of works include research articles, news, letters, and other 
writings published in more than 200 of the oldest leading academic journals. The works date from the 
mid-seventeenth to the early twentieth centuries. 

We encourage people to read and share the Early Journal Content openly and to tell others that this 
resource exists. People may post this content online or redistribute in any way for non-commercial 

Read more about Early Journal Content at 
iournal-content . 

JSTOR is a digital library of academic journals, books, and primary source objects. JSTOR helps people 
discover, use, and build upon a wide range of content through a powerful research and teaching 
platform, and preserves this content for future generations. JSTOR is part of ITHAKA, a not-for-profit 
organization that also includes Ithaka S+R and Portico. For more information about JSTOR, please 




Recent scholarship is surely tending to the view that Epistola x and 
the Quaestio de Aqua et Terra are genuine works of Dante. I need not 
repeat the arguments set forth by Moore, Shadwell, White, Boffito, 
Biagi, and others ; I wish to confirm them, so far as may be, by an 
appeal to the Concordance of Dante's Latin works which Dr. Wilkins 
and I are about to publish. Further, I should like to show that the evi- 
dence there accessible fits better with the theory that the De Vulgari 
Eloquentia (= V.E.), Epistola x (= Ep. x), Quaestio de Aqua et Terra 
(= A.T.), De Monarchia (= J/.), were written in the order indicated 
than in any other order. On different grounds, other scholars have 
reached the same conclusion. For instance, — I am attempting no elabo- 
rate bibliography here, — Wicksteed and Howell ^ assign V.E. to the year 
1304; Ep. X is placed by different scholars between the years 13 16 and 
1319;^ A.T. must have been written shortly after the subject of the 
treatise was discussed by the author at Verona, January 20, 1320, which 
date the work itself gives us. The date of M. is the most mooted of all. 
Some put it fairly early, as Wicksteed and Howell* who decide tenta- 
tively for 1309, but others* regard it as a work of the last years of 
Dante^s life (1318-1321). One bit of evidence on which the last-named 
theory depends may be corroborated, I believe, by the Concordance. 

'^ A Translation of the Latin Works of Dante, 1904, p. 119. See also Paget 
Toynbee, A Dictionary of Proper Names and Notable Matter in the Wbrks of 
Dante, Oxford, 1898, p. 214. 

2 See Mqore, Studies in Dante^ Third Series, 1903, p. 345. 

8 Op. cit. p. 289. 

* See Scartazzini, Enciclopedia dantesca^ 1898-1899, vol. ii, p. 1268, and also 
various references in Pasquale Villari, // "Z>^ Monarchia " di Dante Alighieri in 
Nuova Antologia, vol. xlvi (1911), pp. 393 ff- Villari himself thinks books i and ii 
were written in 1300 and book iii after 1308. There is no support for this theory in 
the evidence which I offer in the present article. 



In i?/] i. 12. 42, Dante apparently makes a direct reference to the 
Paradiso, He remarks : 

Hoc viso, iterum manifestum esse potest, quod haec libertas, sive principium 
hoc totius libertatis nostrae, est maximum donum humanae naturae a Deo col- 
latum, sicMt in Paradiso Comediae iam dixi ; quia per ipsum hic f elicitamur, 
ut homines, per ipsum alibi felicitamur, ut Dii. 

So read Witte^s manuscripts, save that in two of them — P (saec. XIV) 
and F (saec. XV) — lacunae are indicated, showing apparently that part of 
the reference to the Paradiso was erased, either by the writers of these 
manuscripts or, perhaps, by those of their originals. F has sicut in , , . 
quia ; P has sicut . . . commedie iam dixi^ quia, 

Now Marsilio Ficino in his translation of 1467 has nothing at all for 
the phrase sicut . . . dixi^ and the early editions, of which the earliest 
appeared in 1559, have not the clause. Witte thinks that merely the 
words sicut dixi are genuine, and he is foUowed by Moore. The real 
reference, Witte states, is not to the Paradiso, but to the beginning of 
this very chapter of the Monarchia, where Dante has dedaxed primum 
principium nostrae libertatis est libertas arbitrii. But beyond the repeti- 
tion of these words which state the proposition proved in the first part 
of the chapter, there is no reference to it in the later sentence. Nothing 
has been said before about ** the greatest gift conferred by God on man," 
to which sentiment the sicut dixi applies. If now we turn to Paradiso v. 
19 ff., we find an unmistakable connection. 

Lo maggior don, che Dio per sua larghezza 

Fesse creando, ed alla sua bontate 

Piii conformato, e quel ch' ei piu apprezza, 
Fu della volontk la libertate, 

Di che le creature intelligenti 

E tutte e sole furo e son dotate. 

This much is cited by Witte. Perhaps we may go a bit further and see 
in the concluding clause of the Latin passage, quia . . . alibi felicitamur, 
ut Dii, an allusion to the last part of the same canto, where Beatrice and 
Dante come upon a thousand radiant beings replete with divine love, and 
Beatrice tells him to " Speak, speak securely and trust even as to godsJ' 
If Dante has not in mind this passage in the Paradiso, which is alto- 
gether apposite, his reference is most puzzling; for there is no other 
passage which is apposite, certainly none in M. 


If Witte wanted to omit anything, he should have cast out the entire 
clause, sicut . . . dixi^ as Ficino and the early editors did. But why did 
they ? Not necessarily because prompted thereto by critical or hyper- 
critical acumen. They both might well have had manuscripts in which 
some such mutilation had occurred as we find in F and P. Unable 
to make any meaning out of the remnant of the clause, they left it 
out entirely. That would be a critical procedure quite in keeping with 
the practice of those times. That the omission in the manuscripts was 
due to accident rather than hypercriticism is shown by its fragmentary 
nature. I submit, therefore, that it is incumbent upon us to accept this 
reference at its face value until it has been absolutely proved worthless. 
That, however, is not the case. I will not deny that the problem needs 
further investigation and that, in particular, the relation of the different 
manuscripts to one another should be fixed. As F and P seem clearly 
related elsewhere, I am tempted to trace their different mutilations of the 
sicut . . . dixi clause to an original mutilation, or obscurity, in their common 
archetype, from which the manuscripts used for the editio princeps and 
likewise for Ficino's translation descended. But, again, this part of the 
subject demands fresh treatment. 

Meanwhile, let us appeal to the Concordance to see if there is any 
reason why Dante could not have said sicut in Paradiso Comediae iam 
dixi, We find, first of all, that he uses a sicut clause of this sort not in- 
frequently in all the four works {V.E., Ep. x, A.T., M.), when referring 
to the works of another writer. Thus : 

sicut dicit Thomas in tertio suo contra Gentiles (J/. ii. 4. 5) ; sicut dixit 
Philosophus in secundo Metaphysicorum {Ep. x. loi). 

Further, he uses the phrase in referring to his own works, or at least to 
the work in which the phrase appears : 

Sicut inferius' ostendimus {V.E. \. 8. 24); sicut inferius ostendimus {V.E. i. 
12. 55); sicut dictum est (K.^. i. 14. 21); sicut in superioribus est peractum 
{M. iii. 2. 2). 

Now I will not deny that an interpolator might notice this habit of 
Dante's and observe it in his own interpolation, or that, unconscious 
of Dante's usage, he might have happened to adopt it. But I am spe- 
cially interested to note that the phrase is sicut . . . dixi^ and not the 
plural, sicut . . . diximus. If it were the latter, there would be a distinct 


probability that the words are spurious. Howso, one may ask, when we 
have just observed the plural in two passages from the F,E.? The facts 
are as follows, and they may be seen in the Concordance under nosy ego^ 
noster, and meus. 

In V.E. the word ego occurs just once, and does not refer to Dante. 
He uses it in an illustration : ut : Figet me cunctis pietate maiore, etc. (^V.E. 
ii. 6. 36). But Dante tells us about himself not infrequently in V.E. He 
uses for the purpose the plural nos, which occurs thirty-eight times. One 
of these instances is a quotation. In some of the others the word has a 
general sense, '' we men in general," and sometimes includes both the 
author and his readers, whom he has invited to join him in an imaginary 
hunt f or the vulgare illustre. But in about twenty cases it ref ers clearly to 
Dante himself. A specially good illustration is Nos cui mundus est patria 
(i. 6. 1 7), where the plural nos is followed by the singular relative. The 
same holds true for noster, while meus does not occur at all. The same 
holds true for dico, for which he always has dicimus. Other verbs in the 
first person are plural, though there may be an exception or two besides 
the rather striking one I have noticed : 

Nec dubitandum reor modo in eo quod diximus temporum^ sed potius 
opinamur tenendum ( V. E. i. 9. 60). 

Now in all his other works, — and I think that however their order be 
determined, few would object to calling them later than V.E.^ — there is 
only one occurrence of this usage, i.e. Ep. x. 85 : 

Sed zelus gloriae vestrae, quam sitio, nostrum parvipendens (" But zeal 
for your glory [i.e. that of Can Grande], for which I thirst, recking little of 
my own.'''') 

Nos and noster occur elsewhere in Ep. x, A.T.^ and M.^ but always in the 
general, never in the special, sense. Thus : 

Hoc etiam est insinuatum nobis in Matthaeo (" We are given to understand," 
Ep. X. 548) ; in die Solis . . . quem . . . Salvator . . . nobis innuit venerandum 
(A.T. 24. 19); licet ostensa sit nobis haec ab humana ratione, quae per philo- 
sophus nobis innotuit {M. iii. 16. 65, 66). 

When Dante wishes to say " I " in his later works, he used ego^ though 
that word and meus are used very rarely in all of them, apart from quo- 
tations, which of course do not concem us here. So though I have not 
proved that Dante must have written the sicut . . . dixi clause, I could at 


least congratulate an interpolator on not having said sicut . . . diximus, 
Surely the burden of proof lies altogether on those who would expunge the 
words from the text, and as no real proof has appeared, we have a right to 
draw from the words the chronological inferences which they contain. 

I will assume, then, that the Monarchia was written in Dante^s later 
period, at least after the fifth canto of the Paradiso was written. Other 
scholars have placed it there on other grounds, particularly for various 
connections with the subject matter of the Paradiso\ if it is put as late 
as this, nothing compels us to place it before rather than after A.T, I 
am inclined to place it after, for a reason that will later appear, and thus 
to regard it is the last of all Dante's works. 

Supposing, then, as others have done, that V.E,^ Ep. x, A.T, and M. 
were written in the order in which I have named them, I will appeal to 
various stylistic peculiarities not to prove this order, but to show at least 
that it is plausible. Proof is impossible. An array of peculiarities common 
to A,T and M, does not necessarily prove that the two treatises were 
written in the same period ; these peculiarities may depend merely on 
the nature of the subject treated. But it is at least possible that similar 
habits of mind resulting in similar traits of expression were not far sep- 
arated in point of time. I think I can show, that, granting the chrono- 
logical order assumed by others, the various stylistic evidence makes for 
that order rather than any other ; and whatever value this point of the 
discussion may have, I am confident that the material soon to be acces- 
sible in the Concordance will establish beyond cavil the weighty arguments 
already adduced for the genuineness of Ep. x and A.T. 

One point deserves special emphasis at the outset. A.T. was first 
published by Moncetti in 1508; the manuscript which he professed to 
have used is not extant to-day. The author of the work both at the be- 
ginning and at the end declares himself as Dante. If the work is really 
spurious, we are concemed with a deliberate forgery, not merely a case 
of mistaken attribution on the part of either the original scribe or Mon- 
cetti. If, then, the work is a forgery, it is more probable that Moncetti is 
the guilty person than that in his innocence he happened to find what 
some one else had forged either in his own day or at some earlier period. 
I will not deny the possibility of such a circumstance, but it is extremely 
improbable. Our most natural dilemma is that A.T. is either the fiction 
of Moncetti or a genuine work of Dante. 


The case is exactly the same with Ep. x. That was first published 
by Baruffaldi in 1700, but he surely did not invent it; it is found in its 
entirety in three manuscripts of the fifteenth century, and the introduc- 
tory part (§§ 1-4) is found in two manuscripts of the fifteenth century. 
The first mention of the letter is probably that of Villani at the end of 
the fourteenth or beginning of the fifteenth century.^ Now in the salu- 
tation of the letter the author purports to be Dante, and even granting 
that this salutation might be a later addition, a writer who gives an elab- 
orate description of the third part of a poem of his which he calls Cantica 
tertia Comoediae Dantis^ quae dicitur Paradisus (1. 257) evidently either 
is or assumes to be Dante. So, as with A,T.^ the hypothesis of mistaken 
attribution must be ruled out at the start. Either this letter is aforgery 
by Villani or some other writer of the fourteenth century, or it is the 
authentic work of Dante.^ 

By " stylistic evidence " I mean not merely coincidence in noticeable 
phrases. For instance, in M, (i. 14. 78) we have the phrase cum Deus 
velit quod melius est, and in A.T, (13. 39) cum Deus et natura semper 
faciat et velit quod melius est, the combination Deus et natura occurring 
also in several places in M. A coincidence of this sort is interesting, but 
not a proof of authorship, since the phrase is just what an imitator, seek- 
ing to give his forgery verisimilitude, would notice. We must find, if 
possible, indubitable traits of a minor nature which no imitator could 
detect, and which therefore bespeak the genuineness of the work. Again, 
I will not say prove. Stylistics and statistics taken alone must be handled 
with the utmost caution. Added, however, to the other varieties of evi- 
dence in our problem, they come as near to certainty as human methods 
can. Approaches to an investigation of this sort have been given by 
Moore,^ and by Biagi in his recent and very elaborate edition of A.T. 
(1907), which he concludes with a word-index, apparently complete ex- 
cept for some of the minor words, giving parallel passages or phrases in 
the other works of Dante ; in this undertaking he found the concordance 
of Fiammazzo valuable. The result shows, he declares, !a perfetta identita 

1 For the above facts, see Boffito's edition, 1907, pp. i ff. 

2 I am neglecting, as I think I have a right to neglect in the present discussion, 
theories of a composite origin of Ep. x. See Moore, J^tudies in Dante^ Third 
Series, pp. 347 ff. 

8 Studies in Dante^ Second Series, 1899, pp. 346 ff.; Third Series, pp. 
324 ff. 


lessicale e sintattica between A,T, and the other works. This perfect 
identity may be still more minutely shown. 

I will treat the subject under two heads : i, Dante's vocabulary ; 2, his 
constructions. As a preliminary, we should note that the works under 
discussion differ in length, but that if we take for comparison not the 
entire M. or V,E,^ but the books of these works, the units in question 
will be more nearly commensurable. Thus M. i has ten pages of Moore's 
edition ; ii, ten and a half ; iii, thirteen and a half ; F.E. i^ eleven and a 
half ; ii, ten and a half ; Ep. x, sbc and a half ; A.^. eight. It will be seen 
that even when so considered we have smaller amounts of text in Ep. x 
and A.T. than in a book of M. or V.E. and hence should not expect so 
many instances of any given peculiarity. We have enough material for 
profitabfe investigation, however. The pages in Moore are closely printed. 
In the editio princeps^ in which likewise no space is lost, A. T. takes up 
thirteen and a half pages. I will not always apportion my statistics to 
the different books, but the reader should keep these general proportions 
in mind. 

Coming first to the vocabulary of the Latin works, we must remember 
above all that a genius like Dante is bound to vary his phraseology. If 
it were true that all the words in A.T. and Ep. x occurred elsewhere in 
Dante, that would indicate the spuriousness rather than the genuineness 
of those treatises. Let us begin by testing as specimens the Latin words, 
whatever their nature, that begin with A, and note those that are found 
only in some one of the different worksl I do not consider here the 
Latin quotations in Dante's Italian works. The number of occurrences, 
if greater than one, is indicated in the parentheses after each word. 

abscondo^ adusque^ alveolus^ Aonius^ astricola, attritus = 6. 


acernus^ Achaemenides^ Acis^ Adria^ Aemilis^ Aetna^ Aetnaeus^ Aetnicus^ 
agrestis^ Aiphesiboeus^ alumnus, ango^ annosus (2), aridus^ arrideo^ arundo, 
arundineus^ arvum^ avidus = 19. 

anhelus (2), armentum (2). Total for Ecl. = 27. 


M. I 

abeo^ abstractum^ acceptabilior (2), acceptabilissimus^ actuo (2), acuo^ ado- 
lescentia^ aegroto^ agibilis^ agito, albedo^ algor, amplior{2\ analytice^ ancillor^ 
apprehendo (2), apprehensivus^ Arago^ aristocraticus, assequor{2\ auxiliatiOy 
Averrois = 22. 

M. II 

abrumpo, Abydos, ^ acerbe^ ^acies^ adoptio, ^adveho,^ adversor^ AeacideSy 
Aeneis (2), ^aes^^ aestivus^ ^aethereus^^ Afri^ Africa (6), Africanus (2), agon^ 
agonista^ Albanus (3), altrinsecus^ ^amabilis^^ amissio^ Anchises^ ancile (2), 
Andromache^ anhelo^ Antaeus (2), ante adv., approbo^ aptus^ arbiter^ ^ ar- 
dentior^ artificiose^ Assaracus, assentio^ Assyrius^ ^asto^ Atalanta^ athleta{6\ 
athletizo (3), athlotheta^ ^Atlantis^^ ^auratus^^ Ausonia^ ^Ausonius^"* avia^ 
avus (2) = 46. 


abnego, absolute (3), absumo^ abundanter^ abundantia^ accidentalis^ ^ac- 
cuso^ acquiescOy adeptio, adhibeo, advoco (3), aequivalentia^ aequivaleo (4)^ 
afficiOy ^affluens^ Agatho, alienatio^ alieno (2), altar^ altercatio, annihilo^ 
apostema, applico, archipresbyter^ architectus^ areola^ Asianus, ^attexo^ 
auctorizo (6), auditio, ^azyma ' = 31. 

acquiro (7), aestus (2), annexus (2). 

activus (2), ambitus (7), assimilo (3), aureus (2). 

adiuvo (5), artifex (4). Total for J/-= 108. 

V.E. I 

abmotim^ accentuo^ accola^ acerbitas (2), adiectio^ adinvenio (2), admoveOy 
Aduaticus{2\ adultus^ advena{2\ advenio, adverbium, aediJicatio{2\aequator, 
aequo^ aetas^ affirmo (4), Alamania, Alamannus, Alexandria, allego^ allu- 
besco^ alteratio^ alterno^ altriplex^ Alvernia^ ambages, amentia^ amicabilisy 
amoenior^ amoveo^ amussis^ Anconitanus (3), Anglia^ Anglicus (2), angU' 
lus (2), anterioritas^ antiquior^ apocopo^ Apulia^ Apulus (5), Aquileiensis (2), 
Aragonia^ architector n., architector vb., argumentor {2\ Arturus^ aspiro^ as- 
sociOj assuefaciOy assuefOj asy/um, audacter^ augustus, avidissimus^ Azzo = 56. 


V,E, II 

accensio, acutus, additio^ admissio^ aemulor^ Aeneidorum{2\ aequalitas{^\ 
alleviOy alloquor, alterus, angelicus, animalis (2), antecedens adj., aporio, 
appendo, Aquinum^ armonia (2), armonizo (6), Amaldus (5), artificiatus^ 
ascensus (2), asper^ aspiratio, astripetus, aulice^ austeritas = 26. 


accentus (4), amplissimus (2), anterior (2), Arctinus (4), arrogo (2), ^w- 
siduitas (2) = 6. Total for F.^. = 88. 

^Z'. I 
adiaceo, affectuosissimus^ affluentia, Albus, attento = 5. 



abstineo, assiduus. 



^Z'. IV 

EP. V 

* acceptabilis^ affectuosius^ agellus, alba, almus, Alpes^ amplexor^ ancillor^ 
animositas^ anne^ annuo, Argus, arrigo^ assurgo (2), attenuo =15. 


advento^ advolo^ aedificium^ altissime^ amarissime, amens, antiquitas, 
Aprilis, aries, armo^ arrogantia^ atqui^ augustalis, avolo =14. 


accumulo^ Agag (2), aggrego^ Alcides, alimentum, allicio, Amalech (3), 
Amata^ Amos^ ampiexus^ amputatio^ angustissimus^ angustus^ arbor^ area^ 
aresco, ascio^ assevero, avello =19. 


abigo^ abominabilis, abvium^ accuratissime, aestimo^ aetemitas, <Vfigo, 
aggenero^ Alcimus, Ambrosius, ara, aranea^ arca (2, * i '), aspergo n., as- 
tronomus^ auriga = 1 6. 

absolutio, affectuose. 


EF. X 

absolutus^ accuratus^ admirabilitas (2), allecturus^ allegoria (2), allegori- 
cus{2)y alleon^ A(lpha\ amicitia^j, * i '), amodo, anagogicus, analogia^ analogus, 
angustia, Apollo (3), ascensio, attentio^ attentus{2)^ attollo^ auditor{2) = 20. 


accessus, adaequatio (5), adimitor^ aequivoce (2), altior (20), antarcticus^ 
Antepraedicamenta^ apparentia^ appensio., arcticus, astrologus, attractio^ 
attraho =13. 

The above list shows that Ep, x and A,T agree with the accepted 
works in their use of a dozen words or itiore which occur nowhere else 
in Dante. The number of such words varies considerably in the differ- 
ent books of the accepted works ; Ep. x in proportion to its size has more 
than M. i and less than M.\\. A.T. has a sufficient number, though less, 
as we might expect from the subject, than in any book of F.E. or M. 
The longer letters show a high proportion, but Ep. iv, most probably 
genuine, has in its page of text only one word not elsewhere used. 

I now subjoin a list of words which illustrates Dante^s general vocab- 
ulary and which shows that Ep. x and A.T. conform to the accepted 
works in the use of words and senses, whether frequent or rare, in Dante. 
I do not include everything here, but aim especially to show Dante^s use 
of minor words and particles and his technical phraseology — his argu- 
mentative apparatus — though some words here registered do not come 
under this heading. Each word occurs in at least three of the four works 
and sometimes elsewhere. If a word occurs in Ep. x but not in A.T., Ep. x 
is added in parenthesis ; A.T. is added if the word occurs there but not 
in Ep. X ; if no work is mentioned in parenthesis, the word occurs in both 
Ep. X and A.T. In a few cases, e.g. amplius, more exact statements are 
made. If the word is printed in italics, it occurs not more than five times 
in any of the works in which it appears. If it is in black roman type, it 
occurs from six to fif teen times in some one of the works ; if in black italic 
type, sixteen to twenty-five times ; if in capitals, over twenty-five times. 
" Arg. " denotes a logical or argumentative term , ' ^ phil. ' ' a philosophical term. 

Accedo (Ep. x), accipio (arg., Ep. x), actus (A.T.), adduco (arg., Ep. x), 
adhuc (arg., rare ; frequent in St. Thomas Aquinas), agens, ago (arg. and phil.), 
aliqualis (Ep. x; cf. aliqualiter M., V.E.), aliquando^ amplius (arg., rare;, 
M., Epp. iii, vi; frequent in St. Thomas Aquinas), apparet (arg.), appello, 


argumentum, a8sero(Ep. x), coepio (A.T.), communiter, CONSEQUENS {per 
consequens and consequens est in all four), considero, deinde, denique, de- 
structio (arg.), differentia (A.T.), e (rare ; generally in the phrase e converso, 
which occurs in all four works), efficio (A.T.), elementum, eo quod^ equidem 
{A.T.), eyidentia (only in the phrase ad evidentiam ; cf. evidens M., evidenter 
M.,V.E.), existimo (Ep. x), exordium (Ep. x), exttgL^ facilior (E^. x), frustra 
{A,T. \gradus, huiusmodi, ibi, IDEM, ideo, immediate (Ep. x ; cf. immediatus 
M., V.E.), immo^ impossibilis, inferius adv. (A.T.), inffuo (phil.), innuo (arg.), 
inspicio (arg.), INTENDO, intra (Ep. x), intueor (arg., A.T.), invicem 
{Ep. x), ita (rare), magis, manifesto vb., materia, melius adv. (Ep. x), mini- 
mus (A.T.), minus^ MODUS {nullo modo M., V.E., A.T.; per modum with 
genitive in all four works), motor (phil.), motus, multo (A.T.), NATURA, 
naturalis, ne^ NEC, neque (very rare), nihil, nil (rarer than nihil\ Ep. x), 
nonne (rather poetical, Ecl., M. ii, V. E., Epp. v, vi, vii, ix, x), nonnullus (Ep. x), 
nosco (Ep. x), nunc (arg.), numquid, omitto (arg.), oppositum (arg., A.T.), 
O STE N lyO^philosophia, philosophor{A.T.\ plus (very rare ; Ep. x), pono (arg.), 
POST prep., postquam (rare ; Ep. x), y^Mus^ praefatus (A.T.), praemitto (arg., 
Ep. x), praenoto (arg.), praesens, praeter, praeterea (rare ; A.T.), primo {primum 
is very rare, see list for V.E. and M., p. 31), principaliter (Ep. x), PRIN- 
CIPIUM, prior (A.T.), prius, procedo (arg. and phil.), propono incl. propositum 
<arg.), VKOVKWJS, proptereaquod, QUAERO, quaestio (A.T.), qualis (rare), 
quantitas, quantumcumque (Ep. x), quilibet, quidem (Ep. x), QUI QUIDEM, 
quin (rare; M. but none in bk. i, V.E., Epp. v, vi, vii, viii), quippe (E^, x), 
quisquam (rare ; Ep. x), quisque (rare ; Ep. x), quo., quomodo (rare ; Ep. x), 
quoniam (rare; Ep. x), RATIO, recipio (arg. and phil.), recte (A.T.), requiro^ 
RES (phil), respectus {per respectum ad V.E., Ep. x, A.T.), saltem (A.T.), 
satis, scientia, scio (in two thirds of the instances, the gerundive sciendum is 
found; Ep. x and A.T. conform), scribo (introducing quotations), secundo, 
semper, sensus, sermo, seu (rare, see sive\ Ep. x), SICUT, significo (Ep. x), 
similis^ similiter, similitudo, simplex (cf. simplicissimus V.E., simplicitas 
M., V.E.), simpliciter, simul, singulus (A.T.), SIVE (cf. seu\ soleo (A.T.), spe- 
cies (A.T.), speculor (Ep. x), spiritualis (Ep. x), SUB, subiectum, substantia, 
subtiliter (A.T.), sufacienter (A.T.; cf. sufficiens A.T., sufficientia M., V.E.), 
sufficio, super (rare ; super = de is very rare), superius, talis, tam (rare, gener- 
ally tam . . . quam\ tango (arg.), TANTUS, teneo (arg.), totaliter, tractatus, 
tum (rare ; A.T.), tunc (a bit more frequent than tum)^ udique, ulterius^ uni- 
^ersum, universus^ unusquisque (A.T.), usque (A.T.), utrum (A.T.), VEL 
(cf. auf), vere (A.T.), VERITAS, verso vb. (Ep. x), VERO \verum very 
rare; Ecl. ii, M. i, ii, Epp. vi, vii, x), VERUS, via "method" (arg., A.T.), 
virtus =potentia (phil.), volo (arg.). 

Having shown sufficiently by testing Dante^s Latin vocabulary that 
Ep. X and A.T. conform in this regard to accepted works, let me now refer 


to various constructions in which they all manifest a striking similarity. 
I will begin with si, 

Si occurs in each of the seven books between twenty and sixty times. 
The indicative is used in a simple condition, protasis and apodosis, in all 
the books. Conditions contrary to fact, with imperfect and pluperfect 
subjunctive, occur in all the works save Ep, x. A usage not common in 
classical Latin is that of the indicative in the apodosis and the present 
subjunctive in the protasis.^ Thus : 

si contingat peccatum in forma artis, materiae imputandum est {M, ii. 2. 22). 

This present subjunctive is clearly not the less vivid future, but hortatory 
or concessive. The indicative of the apodosis is generally either future, 
or some f orm implying f uture time, like the gerundiye. A favorite formula 
in M. is si with considero, e.g. si enim consideremus unum hominem , . . 
videbimus {M, i. 5. 22). This appears in V,E, also, which work and Ep, x 
have inspicio in the same sense. In M, the apodosis is almost always of 
the nature described. In V,E, and Ep, x the usage is somewhat freer, 
the present indicative appearing oftener than in M, instead of a tense 
implicitly future: e.g. 

si quis autem quaerat . . . respondemus ( F. ^. i. 10. 40); si inspiciamus . . . 
videtur {V.E, i. 12. 15); si ergo accipitur . . . manifestum est {Ep, x, 364); si 
essentia sit intellectiva, virtus tota est unius {Ep, x, 395). 

This construction explains the apparent abnormality of ^.T'. 12. 19 : 

Si igitur aqua erit in A, et habeat transitum . . . movebitur ad B. 

The curious use of the future indicative erit with the present subjunctive 
habeat as a second member of the si clause is a comment on the meaning of 
this subjunctive, a proof that it is not less vivid future. Of the many occur- 
rences of the present subjunctive in conditions, every one is most naturally 
explained as hortatory-concessive. Dante^s substitute for the present sub- 
junctive in less vivid future conditions is a lax use of the imperfect, as : 

Et si quis instaret . . . inutilis est instantia {M. iii. 7. 23); Quod si cuiquam 
. . . videretur indignum, Spiritum sanctum audiat {Ep. x. 35); Si igitur aqua 
moveretur ad B . . . movebuntur {A.T, 12. 45). 

What later forger could have penetrated so deeply into Dante's feeling 
about the conditional subjunctive ? I may add that despite the brevity 

1 There are certain approaches in classical usage. See E. F. Claflin in Classical 
Journal^ 1911» pp- 305-307. 


of Ep, ii, which prevents any extensive application of stylistic evidence in 
its case, the sentence si considerentur . . . lux . . . exoritur (28), speaks 
for its genuineness. 

In ^iV. 7. 43 Dante quotes a sentence from the Vulgate in which si 
takes the indicative in an indirect question, and himself uses this lax con- 
struction in V,E. i. 4. 46, recordetur si numquam dixit^ but not in his later 
works. A coUocation likewise found in V.E.^ but not later, \sputa si. 

In the compounds of si, the same principles are observed. Ac si and 
its equivalents take in classical Latin the present subjunctive in present 
time; it is the same hortatory-concessive subjunctive which in Dante^s 
usage had spread to all clauses with si, and which in Boethius was well 
on its way to this development. In Dante these particles, which are very 
rare, take either the present or imperfect subjunctive, as in M. iii. 15. 34 : 
Quod non sic intelligendum est, ac si Christus . . . non sit dominus ; and 
just below : Velut si aureum sigillum loqueretur. Ac si appears in Ep. x, 
504, and, though differing from the ac si clause just quoted from M., 
shows that the writer understands Dante's real usage : et similis modus 
arguendi est ac si dicerem. 

Nisi is most frequently used elliptically without a verb, as quod esse 
nonpotest, nisi quando, M. i. 8. 27, and is generally preceded by a nega- 
tive. Ep. X and A.T. accord with the other works in this peculiarity. 
When nisi takes a verb, the same constructions are found as for si. Thus, 
to take examples of the hortatory-concessive subjunctive : 

Sed hoc esse non potest . . . nisi sit voluntas una {M. i. 15. 56); diesis esse 
non potest . . . nisi reiteratio unius odae fiat {V.E. ii. lo. 31); non potest esse 
concentrica terrae, nisi terra sit . . . gibbosa {A.T. 13. 9). 

Etsi is not used in V.E. and A.T. It appears in some of the let- 
ters, once in Ep. x with the present indicative in both clauses, and 
several times in M., where either present indicative or present hortatory- 
concessive subjunctive is used. Etiamsi occurs once in V.E.^ with this 
same subjunctive, which is fourid likewise in two of the occurrences in M. 
Thus Ep. X and A:T. agree with the acceptedly genuine works not only 
in conforming to the constructions which Dante frequently employs in 
them, but in avoiding those which he avoids. 

Another significant particle is quod, which occurs over f our hundred and 
fifty times in all, the figures for the works in question being : fifty-eight 


in M, i, fifty-four in ii, ninety-four in iii, fifty-eight in F,£. i, forty-two 
in ii, forty-five in Ep. x, seventy-four m A.T. A curious construction 
appears in V.E., — guod vfith either indicative or subjunctive like ut of 
result. Thus : angelus in illa, et diabolus in illo taliter bperati sunt, quod 
ipsa animalia moverunt organa sua^ F.E.i. 12. ^j. Other proleptic par- 
ticles besides taliter in V.E. are ita^ adeo^ tantus^ in tantum. The same 
peculiarity appears in Ep. x. 528 : intellectus in tantum profundat se in 
ipsum desiderium suum . . . quod memoria sequi non potest. But there is 
no trace of this usage in Dante's latest works. In these, however, quod 
is used a f ew times like ut of purpose, thus : oportet quod reducantur ad 
unum hominem, M. iii. 12. 11. Dato quod, Hoc supponatur quod^ ad hoc 
quod are the other phrases f ound in M. ; the construction develops readily 
from the frequent use of quod in the sense of " that,'* plus a hortatory 
subjunctive. Dante had a model in a sentence from the Vulgate which 
he quotes in M. ii. 8. 63, — hocsolum habemus residui^ quod oculos nostros 
ad Te dirigamus. Naturally the mood used is subjunctive, whereas quod 
in a clause of result takes the indicative in all but two instances, in one of 
which {y.E. i. 15. 56) the subjunctive is the apodosis in a conditional 
sentence, in the other of which {V.E. i. 6. 3) the verb intelligantur is 
perhaps due to an easy scribal error, though Rajna may be right in fol- 
lowing the manuscripts.^ 

Now just as Ep. x conforms to V.E. in the quod of result, so we find 
A.T. agreeing with M. in using a quod of purpose : quod potest fieri per 
unum^ melius est quod fiat per unum quam per plura (14. 34). Likewise 
in Ep. iv. 51, a letter most probably authentic, we find : quod contra 
Rhamnusiae spiculajis patiens te exhortor. Certain instances in V.E. and 
Ep. X which also may belong here will be discussed below (page 23). 

Quod occurs with the familiar causal sense, the usage being regular, 
but in the overwhelming majority of cases it signifies '^ that," and ranges 
through all shades of meaning from the classical use of quod after a verb 
like miror {Ecl. ii. 24), to the freest constructions in indirect discourse. 
The indicative mood is regularly used, but if the statement is doubtful or 
denied, the subjunctive. Thus : 

Sed dicere quod Ecclesia sic abutatur patrimonio . . . est valde inconveniens 
{M. iii. 13. 76); sed dicere quod in excellentissima Italorum curia sit libratum, 

1 Note, however, that the two manuscripts T and G read intelligatur^ which 
might well arise from a misreading of u with superscribed stroke (—un) as «. 


videtur nugatio (F.E. i. 17. 43); potesl etiam probabiliter pstendi, quod aqua 
non habeat gibbum {A.T. 13. 32); Credunt enim vulgares . . . quod aqua 
ascendat (A.T 23. 43). 

In two of these instances, it will be noticed, the ^uod clause precedes. 
Dante has a further practice of using the subjunctive if the ^uoti clause 
precedes, whatever the nature of the statement. This rule is abundantly 
exemplified in all the works. Thus : 

Quod autem Monarcha potissime se habeat ad operationem iustitiae, quis 
dubitat? (M. i. 11. 141); Quod autem verum sit . . . sic declaro (Af. iii. 2. 29); 
Et quod unum fuerit a principio confusionis . . . apparet ( F. -£". i. 9. 14); Quod 
autem de divina luce plus recipiat, potest probari per duo (Ep. x. 453); Quod 
etiam sequatur ipsum substare . . . sic declaro (A.T. 16. 44). 

The reason for this peculiarity is probably that by placing the ^uo^i dause 
first, its substance is made a kind of subject for debate, just as a ^uo£i 
clause with the subjunctive is frequently used as the title of a chapter. 
A question is asked, and the prevailing atmosphere is one of uncertainty, 
which does not clear till the positive statement is given in the main verb. 
But put the main verb first and the situation changes ; the feeling is one 
of assurance and the ^uod clause f oUowing expresses that feeling by the 
indicative. An exception which proves the rule is V.E. ii. 8. 80 : Quo^ 
autem dicimus tragica coniugatio^ est quia, etc. (Cf. M. iii. 5. 13.) Here, 
apart from the semicausal f orce of quod, there is no possible doubt of the 
truth of the statement, and the indicative naturally appears. 

Naturally the subjunctive may appear with guod when necessitated by 
other constructions, as when the verb is also the apodosis of a conditional 
sentence contrary tofact(F:^. i. 9. 67 ; 13. 48 ; Ep. bc. 17 ; A.T. 10. 1,5, 7). 
So, too, a hortatory or concessive subjunctive occurs, — a point which I 
have already discussed and may further illustrate by contrasting the two 
sentences following: 

sicut ad hoc: Quod nemo . . . absque fide salvari potest(J/. ii. 8. 28); sicut 
ad hoc : Quod homo pro salute patriae seipsum exponat {M. ii. 8. 11). 

Exponat in the last sentence is the equivalent of exponere debet. It is an 
excessive feeling of the categorical imperative that results-in the state- 
ment : videtur quod quisque versificator debeat ipsum [sc. vulgare illustre\ 
uti (V.E. ii. I. 20), where either utatur or debet would suffice. An inter- 
esting case is A.T. 16. 19-20: dicamus quod non distet ; et ponamus 


quod . . . distet, where the hortatory force in the subjunctive of the main 
verb flows over, unnecessarily, into the subordinate verbs. A bit looser 
still is M, iii. 2. lo : Haec . . . veritas praefigatur^ scilicet quod illud . . . 
Deus nolit. 

We have noticed that the present subjunctive with si is hortatoiy- 
concessive. Such a subjunctive influences that of quod in the sentences 
f oUowing : 

Et si obiciatur de serpente loquente . . . vel de asina . . . quod locuti sint . . . 
respondemus {V.E. i. 2. 45). Hoc . . . attendendum est . . . quod si eptasylla- 
bum interseratur in primo pede . . . eundem resumat in altero ( V. E. ii. 1 2. 74). 

In the first of these examples there is also something of the flavor of 
indirect discourse. A similar and still more natural subjunctive by attrac- 
tion appears when the main verb is apodosis of a conditional sentence 
contraiy to fact, where the second subjunctive has more justification than 
in the preceding instances. Thus : 

sequeretur . . . quod alterum scilicet esset frustra {M. ii. 6. 28 ; so i. 3. 43 ; 
iii. 6. 5 ; 10. 95); iam videretur quod Deus locutus exstkisset {V.E. i. 4. 47); 
unde sequeretur . . . quod terra undique esset circumfusa {A.T. 16. 14). 

The main verb in the subjunctive with a subjunctive in the quod clause 
occurs only in the instances I have quoted. In A.T. we find two cases 
of an antecedent subjunctive with an indicative in the quod clause : 

Manifestum sit omnibus vobis quod, existente me Mantuae, quaestio quaedam 
exorta est (i. i); Et praesciatur hoc, quod aqua non potest esse concentrica 
terrae (13. 8). 

This is most natural ; the categorical nature of the statement in the quod 
clause is so obvious, that an intruding subjunctive is not allowed. Re- 
membering, however, Fi^. ii. 12. 74, shall we say that the usage here is 
not Dante^s ? That were dangerous, especially as exact parallels may be 
found in Ep. vi. 57, and vii. 77, letters admittedly genuine. In Ep. vi. 180, 
the antecedent phrase vestris animis infigenda supersunt has exactly the 
force of praesciatur in the A. T. passage, and is followed by quod with 
the indicative. Again, these are the exceptions which prove the rule, and 
argue much more for the genuineness than for the spuriousness of A.T. 
I may now add that in one sentence in M.^ although the main verb is not 
subjunctive, the indicative of the quod clause is preceded by a dependent 


hoitatory subjunctive, which, as in the example from the A,T.^ fails to 
influence the mood of the following verb : 

Dico ergo quod licet Luna non habeat lucem abundanter, nisi ut a Sole 
recipit, non propter hoc sequitur, quod, etc. {M, iii. 4. 130). 

I have thus far shown that Dante, though not conforming to classical 
usage exactly, always means something by his subjunctives ; in fact he 
uses them subtly. There remains a curious usage in which a certain 
amount of fluctuation appears. The last quotation, completed, reads : non 
propter hoc sequitur^ quod ipsa Luna sit a Sole, At first one might account 
for the subjunctive by the preceding negative, as in M. iii. 6. 39 ; 8. 70 ; 
and ^. 71 2 3 . 5 5 : non propter hoc est necesse quod imitetur. But the subj unc- 
tive is also found frequently after an affirmative form of sequor. Thus : 

Ex quo sequitur, quod . . . Monarchia sit necessaria {M, i. 13. 69). Et ex 
hac conclusione sequitur . . . quod terra aequaliter . . . distet . . . et quod sit 
substans (^.7. 16. 7-10). Compare also J/. i. 11. 88; ii. 2. 44; 7. 17; A.T. 
12. 57. 

There may be a touch of Dante^s favorite categorical imperative in this 
subjunctive ; or it may be that in stating the conclusion of an argument 
he has in mind the subjunctive quod clause in which the original proposi- 
tion might appear at the head of a chapter — a construction which may, 
as we have seen, explain the subjunctive in a quod clause preceding 
the verb. At any rate, the same usage appears after other expressions, 
besides sequitur, which indicate the drawing of a conclusion. Before 
tuming to these, I wish to point out that the indicative is also used after 
sequor. The most striking instance is M. ii. 2. 47, 48, where the sub- 
junctive has just been used : 

Et . . . sequitur ulterius quod divina voluntas sit ipsum ius. Et iterum ex 
hoc sequitur quod ius . . . nihil est aliud quam similitudo divinae voluntatis. 
So M. i. 12. 93 ; 14. 17 ; iii. 2. 48. 

May we not explain this difference thus, — that if he is thinking primarily 
of the process of drawing a conclusion, he uses the subjunctive, whereas 
if his attention is centered on the fact that he has proved, he uses the 
indicative? The essence of the matter may be further illustrated by a 
passage in ^.T^. 6. 8-9 : 

quare oppositum eius ex quo sequebatur est verum, scilicet quod aqua sit altior 
terra. Consequentia probatur per hoc, quod aqua naturaliter fertur deorsum. 



The first quod clause with the subjunctive states a conclusion ; the second 
quod clause with the indicative states a premise, an established fact, of 
use in drawing a conclusion. 

But to tum to other formulae. Consequens est is evidently a synonym 
for sequitur, It takes quod with the subjunctive in J/] i. ii. 139 ; iii. 16. 
8 ; ^jff. i. 4. 41 ; Ep, x. 106 ; ^.7! 15. 15 ; 21. 34. Rationabile est, or 
videtur esse, has much the same meaning, particularly as we find the 
phrase in close connection with consequens est (KE.i, 4. 37-41). It takes 
the subjunctive in the passage just cited, in KE, i. 15. 29, SindA.T. 7. 5. 
Restat means not " it remains to prove " but *' it follows," in J/] ii. 2. 28, 
32, where it takes the subjunctive. Just so re/inquitur, 12. 13, 
A.T. 20. 44, and especially 4. 7-9 : 

Et cum locus tanto sit nobilior [this amounts to a premise] . . . relinquitur, quod 
locus aquae sit altior loco terrae, et per consequens quod aqua sit altior terra. 

Colligitur has the subjunctive with quod {M. iii. 15. 52) and also the 
indicative (M. i. 13. 33). Undejit quod is surely a phrase denoting infer- 
ence ; it is found only in J/., where it takes now the subjunctive (ii. 2. 50 ; 
iii. 3. 26 ; 16. 109), now the indicative (i. 13. 7 ; 15.13). Hinc est quod 
has the subjunctive {V.E.i. 18. 39), or the indicative {M. i. 4. 19 ; 12.27). 
Inde est quod occurs only in Ep. x. 479, 618, where it has the indicative. 
Signum (est) quod is a peculiarity of KE.y where once it has the sub- 
junctive (i. 8. 45), and once the indicative (ii. 5. 34). The usage of F.E. 
goes rather far in allowing the subjunctive after apparet (ii. i. 31) or 
videtur (ii. i. 20) (see above on the hortatory subjunctive, page 19),* and 
just so Ep. X in Propter quod patet quodW\\h the subjunctive. In M. the 
indicative not infrequently appears where the subjunctive might be ex- 
pected, especially in M.\\. 2. 39 ff., where we find Ex his iam liquet quod 
and the indicative, foUowed by sequitur ulterius quod and the subjunctive, 
and that by Et iterum ex hoc sequitur quod and the indicative (see above, 
page 22). Concludo is a word which on the above principle ought normally 
to take the subjunctive; it appears with quod only in A.T. where once 
(23. 51) it has the subjunctive, and once (23. 17) the indicative. Surely 
this subtle conformity with Dante's usage and no less subtle divergence 
from it in Ep. x and A.T. bespeak the genuineness of these works. On 

1 Perhaps the subjunctive with videtur quod betokens the doubtfulness of the 
statement, as often in St. Thomas, e.g., Sum. Conira Gent. iii. 46, 47. 


finding in A,T, two instances of a pleonastic ^uo^/, which nowhere else 
occurs, I regard the proof not as weakened, but as strengthened. Thus : 

Dico ergo, quod si aqua sit in A, et habeat transitum, quod naturaliter 
movebitur did B (A.T 12. 19, 20; so 21. 32, 34). 

Another detail deserves mention here. In one or two instances we 
note that A.T. and F.E. show a common peculiarity which is seen in 
none of the other works. Thus credo quod^ respondetur quod^ rationabile 
est (or videtur esse) quod, the last phrase with its peculiar subjunctive, are 
found m A.T. and V.E. but not elsewhere. Significant concurrences of 
this sort are especially worth noting in view of the date of publication of 
these two works. Moncetti published A.T. in 1508. Ks V.E., oi which 
only two manuscripts are known to-day, was not published in translation 
and was hardly known till 1529, and as the Latin text did not appear 
till 1577, it is not likely that Moncetti was acquainted with the work. 
Noting then the minute agreements between V.E. and A.T. which I have 
indicated, and others which will later appear, we must abandon once for 
all the supposition that Moncetti forged the A.T. Moreover, it is at least 
doubtful whether Moncetti knew M., for although Marsilio Ficino had 
translated the work at the end of the fif teenth century , the editio princeps 
did not appear till 1559. If Moncetti did not know M.^ we must add to 
the coincidences between A.T. and V.E. a. vastly more numerous array, 
as we shall see, of coincidences between A.T. and M. which could not 
possibly have arisen by chance.^ 

Not much chronological evidence may be found in the quod construc- 
tions. We have noted that the use of quod and the indicative in a result 
clause is rather frequent in V.E.j but appears only once later, in Ep. x. 
Another characteristic of VE. is the use of secundum quod, which occurs 
eight times in Book i and eight times in Book ii, and only scatteringly 
in the later works. 

I subjoin a list of the different words followed by quod in the sense 
of " that." 

V.E., Ep. X, A.T^ and J/.; consequens est, dico, manifestum est, patet, 
probo, scio. 

V.E. and A.T.\ credo, rationabile est (or esse videtur), respondetur. 

1 Only the Convivio had been printed when A. T. appeared, as Moore remarks, 
Studies in Dante, Second Series, p. 307. Of course it should not be forgotten that 
M. was accessible in a fairly large number of manuscripts. 


V.E., A.T.y and M.-, apparet. 

V. E. and M. ; hinc est, testor(also inM. testis est, testimonium perhibet), video. 

V.E. alone; allego, argumentor, attendo, considero, ecce, fateor, indagor, 
obicitur, palatur, praetereo, in mente premo, in promptu est, signum est. 

Ep. X, A.T.j and M.\ praenoto. 

Ep. X and M. ; constat, oportet. 

Ep. X alone ; inde est, praenuncio. 

A.T. and M.; arguo, declaro, ostendo, relinquitur, sequitur, scilicet. 

A.T. alone; concludo, imaginor, necesse est, praescio, pono. 

M. alone ; adverto, ait, assero, canto, comprobo, coUigo, dato, deprehendo, 
dubito, habeo, innotesco, liquet, memini, non obstante, praefigo, satis persua- 
sum est, planum est, restat, revelatum est, scribo, suppono, unde est, unde 
fit, vaticinor. 

Qma is used less frequently (one hundred sixty-nine times) in Dante 
than ^uod (four hundred fifty-three times). It also differs in meaning. 
From Furg. iii. 37 : Stafe contente, humana gente, al quia (= ro ort, " simple 
fact ") ^ one might imagine that Dante generally used quia in the sense 
of " that " and not " because." The reverse is true. It means " that " 
in only nine instances. Thus : 

Satis igitur declarata subadsumpta principali, patet quia conclusio certa est 
{M. i. II. 147; so V.E. ii. 10. i, 2). 

In M. iii. 6. 19, it is plainly used for variety or clearhess, as a quod " that " 
immediately precedes in the same clause. M. iii. 9. 132 is a quotation 
from the Vulgate ; iii. 9. 75 and 117 are virtually quotations. In V.E. i. 
2. 31 a quod " that " clause precedes. In V.E.\. 18. 18 (^Quia vero auli- 
cum nominamus, illud causa est), the particle is semicausal, as in Ep. x. 
94 {Cuius ratio est quid). Thus quia "that" is practically excluded by 
Dante, nor does St. Thomas use the particle often in this sense.^ A 
hasty glance at Albertus Magnus and Duns Scotus indicates that their 
practice is similar. Can it be that Dante uses quia in the Purgatorio not 
as meaning " that," a symbol of mere fact, but as meaning " since," and 
suggesting an appeal to some assured principle which the logician employs 
in drawing a conclusion ? Thus : 

Verum quia omnis veritas, quae non est principium, ex veritate alicuius prin- 
cipii fit manifesta ; necesse est, etc. {M. i. 2. 16). 

1 Cf. Pope, Moral Essays, i. 99 : In vain the sage, with retrospective eye,| 
Would from th* apparent What conclude the Why. 

2 See L. Schiiss, Thomas-Lexikon^ 1881, p. 285. 


This usage is extremely frequent in the three authors mentioned. Dante^s 
remark might mean, therefore, " Be content, human race, with established 
principles, and spend not too much time in seeking new truth by ratioci- 
nation." But lest this explanation be thought morq subtle than scholas- 
ticism itself, I would rather take quia in the usual fashion, since it does 
occur, even though very rarely, in the sense of '^ that," and since Dante 
needs a rime-word here. It would be interesting to know whether any 
author of the period used the word regularly in indirect discourse. 

As to other uses of quia^ I will note merely that in all the four works 
the particle not infrequently is initial, with the force of nam : and that in 
jill the verb is now and then omitted, as : 

Cum ergo Monarcha sit universalissima causa inter mortales, ut . . . bene 
vivant, quia principes alii per illum, ut dictum est {M. i. 11. 138); non est 
extra materiam naturalem, quia inter ens mobile {A.T. 20. 1 1). 

In Ep. X. 221-222, this usage is so extended that the particle almost 
means " to wit " : 

Nam si ad materiam respiciamus, a principio horribilis et foetida est, quia 
Infernus ; in fine prospera . . . quia Paradisus. 

In M. iii. 5. 4, the meaning is surely " to wit " : 

dicentes, quod de femore lacob fluxit figura horum duorum regiminum, quia 
Levi et ludas. 

A study of all the constructions used by Dante in indirect discourse 
would be interesting, but I cannot undertake it here. That there is prob- 
ably no fixed ratio between the use of the infinitive and quod appears in 
the constructions with dico^ which will be found in the Concordance. Also 
apparent in all his works is the fondness for using the direct discourse 
after dico. An interesting combination of a quod clause and an infinitive 
clause occurs va M.\. 14. 17 : 

Sequitur, non solum melius esse fieri per unum . . . sed quod fieri per unum 
est bonum, per plura simpliciter malum. 

This may be matched with Ep. x. 344 : 

ubi dicit se fuisse in primo coelo et quod dicere vult de regno coelesti quid 
quid . . . potuit retinere. 

There should be no semicolon after coelo^ as in Moore^s edition. 


I have seletcted only a few syntactical peculiarities for discussion, but 
the reader of the Concordance will find many other matters to strengthen 
his faith in the authenticity of Ep, x and A.T.hy examining, for instance, 
the articles cum^ dutn, ubi, ut {uti\ licet, quamquam^ quamvis, quando, 
quare with indirect question, and indirect questions in general, the re- 
flexives sui and suus^ and the auxiliary use of the perfect of sum. The 
evidence therein contained, added to what we have already noted, should 
dispose once for all of the possibility of forgery on the part of a later 
writer like Moncetti, who lived at a time when Latin style had undergone 
a thorough transformation. One who would declare A,T, and Ep, x 
spurious is thus driven to the supposition that both treatises are forgeries 
of the fourteenth century. But even then the burden of proof would rest 
upon him : he must explain away the many minute coincidences with 
Dante's genuine writings. Complete evidence cannot be presented until 
further study is made of the writers of Dante^s time, that common traits 
of the period may be distinguished from peculiarities of Dante. It would 
be profitable, for instance, to examine the writers quoted by Biagi^ 
who were interested in the subject discussed in A,T I will appeal to one 
example of an almost contemporary style, a style at least antedating the 
new humanistic Latinity,^ namely that of Villani. The opening sections 
of his commentary on the Inferno are of special interest, seeing that his 
material is based in part on Ep, x. I note certain resemblances to Dante's 
usages ; the use of a present subjunctive in a si clause seems similar.' 
But a rapid glance reveals several important details in which Villani is 
not at one with Dante. One is a frequent use of siquidem, generally post- 
positive, in the sense of enim, which I find in Villani wherever I tum ; 
Dante does not use siquidem (or si quidem) at all. Another striking fact 
is that in Villani quod ^* that " hardly occurs. In the first thirteen chapters 
of the Comento, a section surely larger than Ep. x, I have discovered 
only two after a hasty search, and there cannot be many more. One of 
these is especially interesting. It occurs in chapter x (p. 34 Cugnoni) 
where Villani is defining comedia. He says : 

Ad quorum intelligentiam scire debemus, quod ab hoc greco nomine comos, 
quod latine villa sonat, et oda, cantus dicitur comedia, hoc est villanus cantus. 

1 Op. ciu, pp. 18 1 ff. 

2 See G. Cugnoni's edition of Villani's Comento al Primo Canto delV Infemo^ 
1896, in Passerini, Collez. di Opusc. Dant.y vol. xxxi, pp. 18 ff. 


But this is virtually a quotation of Dante^s words' {Ep, x, 190 ff.): 

Ad cuius notitiam sciendum est, quod comoedia dicitur a contus^ villa^ et 
oda^ quod est cantus, unde comoedia quasi villanus cantus, 

It is curious that the author of Ep, x should use sciendum est, which is 
very frequent in Dante, while Villani should change this to scire debemus 
which Dante never uses. A bit later (/. 218) Dante declares : 

Et per hoc patet, quod Comoedia dicitur praesens opus. 

This time Villani (p. 35) changes the construction to one more familiar 
to him : 

Bene igitur, si diligenter opus totum nostri comici spectetur, rite comedia 

If Villani forged Ep, x, as some believe,^ he possessed both tremendous 
intuition and a most curious method. Rather he is dealing with a source, 
and the personality of the forger, as in the case of ^.T'., must be pushed 
further back, — back, I believe, until it loses its hypothetical existence 
and merges with that of Dante himself. 

But I tum now from syntax to Dante^s vocabulary again, in the hopes 
of finding evidence not only for the genuineness of the disputed works, 
but for the chronological order assumed at the beginning of this paper. 
Let me state again that I mean this as deductive, not inductive, proof ; 
starting with the order V,E.^ Ep, x, A,T,^ M,^ I aim merely to show that 
stylistic traits are in conformity with it. 

Assuming V,E, to be the earliest of the four works, we find the fol- 
lowing words or idioms used exclusively (black letter) or largely (italics) 
in this work, but rarely or never in the three later works ; occurrences 
elsewhere than in the four works are sometimes indicated in parentheses. 
As in the general list given above (p. 1 5 f .) I have excluded words which 
seem primarily demanded by the nature of the subject, though in this 
matter it is hard to draw the line. 

affirmOy attendo, breyissimus, breviuSy ceu^ circa^ comminiscor, conceptio, 
conceptus, conicio, consensus, consequenter, contanter, conyinco (arg.), corporali- 
ter, cunctus^ deinceps^ demum^ diffinio (V. E. ii), directe^ discretive, discussio, 
disiunctim, disiungo, dissentio, dissero, dissuasorie, doctrina^ dumtaxat^ elucido, 
etenim (rare in M., A.T., notin Ep. x), examino, excellens {excellentior Ep. x), 
excellenter, excellentia (A.T.), excellentissime, excellentissimus^ excellentius, 

1 See Moore, Studies in Dante^ Third Series, p. 345. 


ezcello, exinde, extimatio, eztrinsecus adj. {extrinsecus adv. Ep. ii), extror- 
sum, factura, falsissimus (but ci.fallo etc. under A.T. and M., pages 33, 34), 
fateor,y^r^, figurate,\s, forte {conXxz&t forsan, fortasse, yi.^ forsitan 
M., Ep. x), frequentior (cf. frequens Ecl. ii), frequenter {frequentius Ep. iii), 
frequento, gratulanter, habituo, idcirco^ imitatio (cf. imitor M., V.E., A.T., 
imitabilis A.T.), imperfecte (cf. imperfectus M.), incongruus, inconyenienter 
(cf. inconveniens M.), incunctanter, individuum, innovo, ironice, irregularis, 
lector (vocative), membratim, mentio, mixtura {mixtio A.T.), mox^ multimode, 
necnon (very rare in the other works), necubi, nempe^ nequicquam, ;«', nugatio, 
num, omnimode, oretenus, orientaliter, paene^ partim, passim, paulatim, pendo, 
penitus^ penso, perpendo, perplures, perscrutor, perspicaciter\ perspicio, per- 
suasio, persuasorie {persuadeo^ persuasor^ M.), pertracto (once in M., Ep. x), 
pessime, posterus {in posterum V.E., Ep. i), postmodum, praecedenter {prae- 
cedens M., V. E., praecedo M., V. E., A.T.), praeimmediatus, praepono^ praero- 
gativa, praerogo, praetermitto^ primitus, principio vb., progressio, progressive, 
proinde, prorsus^ puta^ puto^ quamplures^ quapropter^ quare (cf. especially 
initial quare in the sense of igitur\ quicquid (elsewhere quidquid. But are 
our texts certain on this point?), quis indef. (distinctly less frequent later), 
quomodocunque, quotquot, raro {rarius Ep. ix), rarissime, ratiocinor, rationa- 
biliter, rationabilius, rationabilis (A.T.), rationalis^ recolo^ reviso *^ to review," 
saepissime, secundarius {secundario M.), segregatim, seligo, sensibilis, sensu- 
alis, sensualitas, significatus, singulatim, spirituatus, subintelligo, subsecun- 
darius, successio, successive, successivus^ supercedo, superexcellentia, taliter, 
tenus (only in superficie tenus V.E. ii), tot^ tottot, tracto, trifarie, trifarius, 
utrinque, utrobique, verumtamen, vestigo, videlicet {scilicet is more frequent 
in M., Ep. X, A.T.), videtur (arg.), vilipendo, voco. 

One interesting peculiarity I reserve for the last, — Dante^s use of venor. 
In V.E. he engages his readers in an imaginary hunt for the vulgare 
illustre. Decentiorem atque illustrem Italiae venemur loquelam, he declares 
{V.E. i. II. 3), and after the search, postquam venati saltus et pascua 
sumus Italiae (i. 16. i). The word is used thus figuratively ten times in 
all, and it is not surprising that later, in J/], he should twice revert to it 
in a general sense : Ad bene quoque venandum veritatem quaesiti (ii. 8. i) 
and hanc veritatem venantes (iii. 3. 113). Clearly these passages are 
later than those in V.E., granting that Dante is the first to use venor 
thus colorlessly. 

The above instance is typical of a certain quality in V.E.^ namely an 
endeavor to substitute the picturesque — sometimes the grandiloquent — 
for the technical terms of argumentation. There is a seasoning of poetry 
in the style of V.E. This peculiarity is not maintained later on, even in 


M, ii, which in theme and in spirit is even nearer to poetry than V.E, is. I 
will quote one more of many possible instances. Contrast Postquam . . . 
deveritate primae . . . inquisitum est, instat nunc (M. ii. 2. 3) with Praepa- 
ratis fustibus . . . nunc fasciandi tempus incumbit {V. E. ii. 8. 2). Other 
instances of this picturesqueness or grandiloquence may be found under 
cribro^ decerpo^ depompo, divarico, extricatus^perplexus^potiono^progressio. 

The above list of favorite expressions of V.E. which occur rarely or 
not at all later may be supplemented by a list of those developed later, 
being found rarely or never in V.E. Various instances have already 
been given in the general list above (page 1 5 f .). To these I would add 
the following: 

arguo {argumentor appears only in V. E.\ manifestus^ sic (fewer in V. E. i 
than later). 

V.E. 3nd Ep. X, according to the hypothesis that I am foUowing, were 
separated by twelve years or mpre. We shall therefore not expect to find 
many significant peculiarities in which they agree against the other works. 
The foUowing, however, may be mentioned : 

Affinitas, alias, alternus, breviter, congruus, ergo (In V.E. and Ep. x ergo 
is almost always postpositive. In A.T. and M. Dante shows a preference for 
initial ergo, especially in A.T. and M. ii. When it is used postpositively in 
A.T. and M., the part preceding is almost always a minor word, e.g. cum, ubi, 
si, est, dico. In the earlier works the usage is freer in this respect; e.g. opor- 
tuit, rationabiliter, praesumpsit, trilingues, etc. in V. E. ; praeferens, differt, 
dividitur, vidit, in Ep. x), generalis, generaliter, hucusque, libet, plerumque, 
postea, praelibo{2iX^.\ praetereo, primordium, que (rare in Dante except in poetry 
or in poetically flavored prose, such as V. E. i [twenty-two times], ii [ten times], 
M. ii [twenty-seven times, of which nineteen are quotations], Ep. vi [seven 
times]. It is not strange that que does not occur in A.T.; in M. iii it occurs 
only twice. Ep. x with eight instances conforms to V. E. ii), quidni, quod with 
the indicative in a clause of result (see above, page 19), quoque (cf. que. 
Found in Ecl., twice ; V. E., fifteen times ; Ep. x, twice ; none in A.T. or M. 
except three in M. ii), recordor, sector (arg.), trado (" set down," "give," deriv- 
ing from the meaning " hand down," which we find in M. : ut Lucas in Evan- 
gelio suo tradit. Cf. V.E. : Volentes igitur modum tradere quo, etc, and 
Ep. x: Volentes igitur introductionem tradere, etc). 

From the above occurrences one could never prove that V.E. and Ep. x 
were written at the same time. They are useful indications, however, 
that the two works are by the same author. 


We have noticed before the significance of coincidences between A,T. 
and V.E, if, as is well-nigh certain, Moncetti was not familiar with the 
latter work. To those already given I may add the foUowing, which, as 
with the list just given, indicate identity of authorship though not adjacent 
dates of composition. 

Accido, aequivocatio, artificialis, cognitio, consimilis, contra (arg.), diversi- 
fico, diversimode, diversitas, ezcellentia (cf. excellens etc. in V.E.), donec^ 
identitas, instruo^ praescio, quasi "almost" (V.E. i. 8. 34 ; cf. A.T. 19. 60, 63), 
rationabiHs, refert, regularis, resulto, stultitia, ultra adv., versus prep. 

The foUowing coincidences between V.E. and M. are worth noticing: 

Abhorreo, absurdus, accidens, actio, adeo adv., adiutorium, aequalis, aliqua- 
liter, antequam, ascisco, astruo, attestor, beneplacitum, converto (arg.), distinguo 
(arg.), dubito, dubius, dummodo, dupliciter, edoceo, elicio (arg.), enucleo (arg.), 
ezpresse, facile adv., gradatim, informo, ingredior, innotesco, intentatus, ite- 
rum, manifeste, medium (arg.), mensura (arg.), mensuro^ minime, modo, neces- 
sario, nemo^ nuncupo, obicio^ obiectum^ paucus^ plerique^ prae^ praeallegatua, 
praesertim, primum^ priusquam, probatio, procul dubio, prorsus^ qualiter, 
quodammodo, rectius, regula, resumo (arg.), singularis, solutio, speculatio, 
statim, subsisto^ testor^ testimonium^ ubicumque, ullus^ unicus^ umquam. 

Once more, while this list shows significant coincidences in minor 
usages which bespeak a common authorship, one could not prove from 
them that the two works were written in close succession. Some of the 
peculiarities are, as noted, found in other works as well, and the number 
of those that remain is no greater proportionally than that given for V.E. 
and A.T.^ which is not one third the length of M. There is thus no con- 
firmation here of Wicksteed^s theory as to the date oi M. \ can add that 
though V.E. i and ii show individual differences, just as the books oi M. 
do, there is nothing to indicate that, as some have supposed, they were 
written in different periods. 

Tuming now to Ep. x, we find just as with its general vocabulary, 
tested by words beginning with A, so with the minor peculiarities in 
question, that the individuality of the author is shown by certain words 
or usages which occur mainly or only here. The list is : 

causo^ circumlocutio, circumloquor, connaturalitas, consideratio, consonanter, 
consuesco, convertibilis (arg.), corporalis, credulitas, definitivus, descriptivus, 
digressivus, dispar, divisivus, doctrinalis, dogma, duplum, excellentior (cf. 
excellens etc. in V.E.), excessivus, excessus, executivus, existentia, exordior, 


expono (arg.), expositio, exterminium, fictiyus, formabilis, formativus, formula, 
improbativus, incertitudo^ inchoo, ^infinitus/ insinuo, investigatio, literalis, 
literaliter, metaphorismus, negotium (phil.), nimis, nullatenus, obyio, percen- 
seo, persaepius, polysemus, positivus, possibilitas, posterius^ praenunciatio, 
'primarius/ probativus, risibilis (phil.), sempitemo vb., sententio vb., seor- 
sim, subtilis, suppositio, transumptiyus, votivus. A fondness for adjectives 
in ivus {tivus) appears in Ep. x, with which only M. may be compared in 
this respect. 

I have noted only few coincidences between Ep, x and A.T, not else- 
where found, and owing to the brevity of these works we should not 
expect many. 

The following are the most significant : 

adaequo, designo, ad evidentiam dicendorum, ethica, magnitudo, materialis. 

For Ep* X and M, there is a much longer list, from which, however, we 
could not infer that the dates of these two works lay in close proximity. 

Allegorice, amplio, antecedens n., aperte, assigno, assumptio (arg.), at, 
causo, compendiose, competit, connecto, defectus, deyenio (arg.), discurro, dispo- 
sitio, doceo, elongo, essentia, exprimo, factum, forsitan (cf. forte etc. in V. E.), 
incorruptibilis, infinitum, intellectiyus, intellectualis^ intelligentia, introitus 
(arg.), intuitus, iuxta, licentio, liquet^ manifestatio, mediate, moralis, mysti- 
cus, nequaquam, notitia, perduco {ad vitam aeternam M., ad statum felici- 
tatis Ep. x), practicus, praeemineo, praeeminentia, praefigo, processus (arg.), 
propterea^ prosequor (arg.), quinimmo, relatiyum, saepe^ salutatio, specialis, 
speculatiyus, suadeo^ subicio (arg.), supra adv. 

I have found no significant evidence whatsoever from stylistic peculiari- 
ties against the genuineness of Ep. x. 

The individual peculiarities of A.T. are : 

certior (cf. certitudo M., A.T.), circiter, citissime, citra^ concupiscibilis^ con- 
fingo, continue, demonstratio (M.), demonstro (M.), disco^ disputo (cf. disputatio 
M., A.T.), exaro, homogeneus, idealiter (cf. idea M.), imaginatio, imaginor, 
imitabilis (cf. imitatio etc. in V. E.), impossibilitas, * incomprehensibilis,' in- 
discussus, indubitabiliter (cf. indubitabilis M.), inductio, innatus, inobedientia 
(phil.), membrum (arg.), miscibilis, mixtio, mobilis (cf. mobile M., A.T.), neuter 
(arg.), obedibilis (phil.), obedio (phil.), ostensiyus (cf. ostensive M.), potentia- 
tus, priyatio, probabiliter (cf. probabilis M.), qualifico, restringo (arg.), sensi- 
tivus, substo, sufficiens, terminabilis, unif ormiter (cf . uniformis M. and A.T.), 
yirtualis {yirtuo^ virtuosius^ only in M. and A.T.). 


It will be noticed that about a fourth of these words appear also or 
have analogues in M. Coming now to a complete, or nearly complete, 
list of coincidences between A,T, and M,^ one cannot fail to be impressed 
by its length and significance. 

admitto (arg.), adverto^ aequaliter, aequinoctialis, alibi, ambo, apud (cf. 
apud negantes divinam bonitatem^ apud oblique politizantes^ M., and apud 
recte philosophantes A.T.), certitudo, circulatio, complexionatus, conclusio, 
conjinno^ consequentia, ^<?^/^r(C?, deorsum, destruo(arg.), determinatio, determino, 
dispensator, dispenso, disputatio, dissolyo (arg.), distinctio (arg.), documentum, 
efficacia (arg.), efficiens (phil.), ens, ex (is rare in V.E. and Ep. x and is used 
only four times in the former, never in the latter, in argumentative phrases, as 
ut ex prae^nissis manifestum est. But in A.T. there are fourteen instances of 
ex in this sense, while the occurrences in the different .books of M. are more 
numerous still. The phrase ex parte * with respect to ' [e.g. ex parte boni . . . 
ex parte vero mali^ is found only in A.T. and M. The phrase ex notioribus 
nobis [used of drawing an inference] is found in A.T. 20. 20 and Ep. v. 122), 
ezcludo (arg.), experientia^ facilis {facile est A.T., de facili M.), facillime, 
fallo^ falsitas, figura {per primam or secundam figuram\ finalis^ fundo vb. 
(arg.), generabilis, genero^ ibidem, includo, influentia, infra adv. {ut infra 
patebit M., A.T.), inquisitio {inquiro in all four works), instantia, insto (arg.), 
insum, insuper, item (arg.), maior (arg.), manifestissimus^ melius est, menda- 
cium (arg.), minor (arg.), mobile (cf. mobilis A.T.), multoties, naturaliter, 
necesse^ notus^ nuUus adj. (" nothing worth," as dico quod sua probatio nulla 
est and et sic . . . instantia nulla estM.-, sed talis instantia nulla est A.T.), 
opinio, opinor, particularis (cf. particulariter, particulo, M.), possibilis, po- 
tentia = SiW/jws {ci. potentiatus A.T.), potissime, potissimus, praedicare (arg.), 
principalis (arg.), prohibeo, proportio, propositio, propter primum {propter 
primam partem Ep. x), quaestio, recipio (phil.; in quantum propria natura 
[natura rei] recipere potest \recipit'] M., A.T.), relinquitur quod, removeo (arg.), 
secundum quid, solvo (arg.), sophisticus, subiaceo (phil.), subiectus (phil.), sub- 
tiliter (cf. subtilis Ep. x, subtilius V.E.), suppono (arg.), syllogismus, tango 
(arg., quod [ut] superius tangebatur M., A.T.), theorema, totalis {totaliter in 
all four), ultra prep., uniformis (cf. uniformiter A.T.), unitas, universalis, 
valde, virtuo {virtuans M., virtuatus A.T.; cf. specificatus M., spirituatus 
V. E.), virtuosius, vis (arg.). 

This is too long a list of coincidences, it seems to me, to explain merely 
by the fact that A.T. and M. are nearer in theme to one another than to 
the other works. Exact statistics in a matter of this sort are impossible, 
but it is safe to say that the number of significant coincidences with M, 
in A.T. is twice that in Ep. x, and twice that in V.E.\ this latter reckon- 
ing, f urther, should be more than doubled since V. E. is more than twice 


as long as A.T, I am tempted, therefore, to regard as the explanation not 
merely the f act that Dante at two different periods argues abstrusely and 
hence falls upon the same terms ; M.\\\s, qiiite as poetical in feeling as 
V.E.^ and yet it conforms in these peculiarities to the style of the other 
books. My theory would be that Dante, impelled by the nature of the 
subject, employed logical terminology more systematically in A.T. than 
he had done before, and that he continued this style and developed it in 
the work that immediately foUowed, namely his last wor^c, M. The mood 
in which he had written A.T. was still on him.^ 

As with Ep. X, I have found no stylistic usages in A.T. which argue 
against the genuineness of that work. 

Let us finally consider the peculiarities of J/, which not unnaturally 
outnumber those of any other work. 

aut (until M. vel is far more frequent), commode, compositio, cotnprobo, 
concorditer, conscribo, consequor (arg.), consonat, consonus (cf. consonanter 
£p. x), constituo, constitutivus, constnio (arg.), contradictorium, contrarietas, 
contrarior (cf. contrarius^ contrarium^ M., V. E., A.T.), cooperatio (phil.), cor- 
ruptiyus, * credibilis,* declaratio (cf. declaro especially in M. and A.T.),definitio, 
destructive (arg., cf . destructio M., Ep. x, A.T., destruo M., A.T.), deviatio, diffe- 
rentialis (cf. differentia M., V.E., A.T.), directivum, directivus, diremtio (arg.), 
dispono, distinctivus (cf. distinctio M., A.T.), distribuo (arg.), distributio (arg.), 
dubitatio (cf. dubito M., V.E., etc), efficax (arg.), efficacissimus (cf. efficacia^ 
efficiens, M., A.T., efficio M., V.E., A.T.), erga, erro (cf. errorU., V.E., Ep. x, 
etc), evidens, evidentissime (cf. evidenter M., V. E., evidentia M., V. E., Ep. x, 
A.T.), executor, expressus (cf. expresse M., V. E.), extensio (phil), extremitas 
(arg.), de facili, facilius, adv. (cf. facile est A.T.), facile adv. (facilior M., 
V.E., Ep. x,facillime M., A.T.), factibilis, /^/y«j, (A.T. etc, ci.fallo M., 
A.T., cX,c.,falsissimusY.^.,falsitas M., A.T., cic), Jiguro {ci.figurate V.E.), 
fingo, finitus (phil.), formale, formaliter, formo, forsan, fortasse (cf . forsitan 
M., Ep. X, tic., fortassis Y.Y^.forte V.E. etc), fundamentalis, fundamentum 
{ci.fundo arg., M., A.T.), habitualis (phil.), habitus = l^t? (used differently in 
V.E.), iam (arg., M., V.E.?), idea (cf. idealiter A.T.), illatio (arg.), immani- 
festus, imperfectus (cf. imperfecte V.E.), importo, imputo, inconveniens, in- 
corruptibilitas (cf. incorruptibilis M., Ep. x), incredibilis, indispositio, indis- 
positus, indubitabilis (cf. indubitabiliter A.T.), inductivus (cf. inductio A.T.), 
infallibilis, infero (arg.), infra prep. (cf. infra adv., A.T. etc), inopinabilis, 
inquam, intentio (A.T. etc), interemptio (arg.), interemptivus, interimo (arg.), 
introduco, irrationabilis, irrefragabilis, iterum (arg., V. E. ?), iuxta, logicalis, 

1 I v/ill not deny the possibility, suggested by Dr. Wilkins, that Dante began 
M. first and wrote it and A. T. at the same time. 


logicus, longe, medio vb., medium (arg.), memini, minoratio, narro (cf. nar- 
ratio Ep. x), necessito (cf. necessitas M., V.E., Ep. x, A.T., etc, necessarius 
M., V. E., A.T., etc, necessario M., V. E.), nefas^ negatio, nihilominus^ nimi- 
etas, nondum^ ob^ oblique (arg., cf. obliquus M., V.E.), obsto (only in nulla vi 
. . . obstante^ non obstante quod\ omnino^ operatio (once in V.E.),,operativus, 
optime^ ordino (V. E., A.T., etc), ostensive (cf. ostensivus A.T.), otiose, otiosus 
{^h\\,\ pariter et^ partialis, particulariter, particulo (cf. particularis M., A.T.), 
patentissimus (arg.), * paulo,' per prius, perago (arg.), perhibeo, persaepe (cf. 
persaepius Ep. x), perseitas, personalis (cf. persona M., Ep. x, A.T.), per- 
sxiajSi^tnXxoty persuadeo^ persuasor {ci. persuasio^ persuasorie^ V. E.), pertinaciter 
(arg.), pertingo (phil.), philosophicus {ci. philosophia Ep. x, A.T., etc), placet (as 
ut Philosopho placet^ cf. placuit A.T. etc), pluralitas, plurimum adv., jporro,' 
post adv., praedicatum (cf. praedico M., A.T., etc), praeoperor, praeostendo, 
praepeditivus, praesentialiter, produco (phil.), productio, profecto, prohibitio, 
prohibitiyus (cf. prohibeo M., A.T.), proprietas, proprius adv., * quamdiu,' 
quatenus^ quousque, realis, rectrix, recurro (arg.), redarguo, reduco (arg., V.E., 
Ep. X, A.T., etc; cf. habere reduci M., A.T.), refello, refuto, regulatrix, re- 
narro, resolvo (arg.), respective, rursus (arg.), sane, scriba (cf. scribo M., Ep. x, 
A.T., etc), secundario (cf. secundarius V.E.), sortior (arg.), specialiter (cf. 
specialis M., Ep. x), specificatus, specto, sponte, stricte (arg.), stultum est, 
suasio (cf. suadeo M., Ep. x, etc), subadsumo, subdo (arg.), subinfero (arg.), 
subito^ subsequens (arg.), substantialis (cf. substantia M., V.E., Ep. x, A.T.), 
superficialiter, superfluitas, supematuralis, syllogisticus, syllogizo (cf. syllo- 
gismus M., Ep. iv, A.T., syllogizator Ep. v), tandem, tantummodo, terminus 
(arg.), testis, testimonium (M., V. E., etc ; cf . testor M., V. E., etc), theologica, 
theologicus, theologus, totidem, typice, typus, ultimo, ultimtmi, ultimus, uni- 
versitas^ universalior, universalis^ universalissimus, universaliter^ utinam, 
vicis, volitivus, volo {velle as noun), voluntarie, voluntas. 

In proportion to its size, there are no more striking evidences of inno- 
vation in M. than in V,E.\ but the innovation is of a different kind. 
Whereas in V.E.^ as we have seen, Dante strives for the unusual and 
picturesque, in M. he starts with the ordinary vocabulary of the logician, 
which, according to my theory, he had just been using in ^.7!, and then 
greatly develops that. Such a development is obvious from the foregoing 
lists, and I may further illustrate it by one striking example, — the uses 
of patet. The frequency of this word, in a variety of phrases, must, as 
Dr. Moore well remarks,^ impress eyery reader of the Latin works of 
Dante. I will try to show also that these phrases form a crescendo. 

Dante wst^ pateo in VE. nine times. It is used either absolutely, or 
govems the infinitive, a quod clause, or an indirect question. It is found 
1 Studies in Dante., Second Series, p. 348. 


in the foUowing set phrases : Etsic patet^ Quare . . . patet^ per quod patet^ 
satis etiam patere videtur^ and with ut ; ut per inferiora patebit, Not to 
notice one occurrence in Ep, iv and one in EcL i, I pass to Ep, x, which 
has seventeen occurrences, a much larger proportion, with the same con- 
structions as in KE. and with set phrases as follows : Et sicpatet (as in 
V.E.), Etper hoc patet^ Propter quod patet^ Sic ergo patet, Patet ergo quo- 
modo^ Potest amodo patere quomodo, and with ut ; ut patet de, ut patet per, 
A dative is also used, as Persaepius inspicienti patebit, A,T, has the 
largest proportion of occurrences, thirty-three in all, with the same con- 
struction as in V,E. and Ep, x. For phrases it has Et sic patet (Ep. x 
and V,E.)y Per quodpatet (as V.E., but nowhere else), sic igitur patet (cf. 
sic ergo patet as Ep. x), et quod and the subjunctive preceding patet. 
Phrases with ut are especially cultivated : ut patet^ ut infra patebit, ut 
patet ad oculum, ut de se patet, ut patet per (as Ep. x), ut patet ex, ut 
patet in, ut patet intuenti(d.. i?ispicienti patebit, Ep. x). In M. the occur- 
rences f or the diff erent books are : i, seventeen times ; ii, twelve times ; 
iii, twenty-one times. This is a less number proportionately than for 
A.T., but the usages are distinctly more varied. Of the constructions 
that have already appeared we find the foUowing: et sic patet, et per hoc 
patet, propter quod patet, sic ergo patet, et quod with the subjunctive pre- 
CQding patet, ut patet, ut patet de^ ut patet per^ ut patet ex, ut patet in, ut 
patet with a dative, ut infra patebit, ut de se patet. Besides these are patet 
quia (as well as quod\ patet igitur quod, ex quo patet, ex iis ergo . . . 
patet^ hinc etiam patet^ et hinc etiam patere potest. To the ut phrases, 
ut statim patebit is added ; sicut patet appears for the first time, likewise 
quod patet^ quod patet de levi^ quod de se patet. Clearly there is a natural 
development, though not a rigid arithmetical progression, in the use of 
pateo from V.E. through M. The case is typical of what I am con- 
vinced is true of the style of M. in general. 

Statistics, I would repeat, and especially statistics of stylistic pecuKari- 
ties, are fraught with danger. They may at least serve as an imperfect 
symbol of the feeling which I have slowly formed about the works in 
question. Such a feeling on the part of the calculator of minutiae is to 
the calculator a most important element in the calculation, though it can- 
not be communicated directly.^ From the evidence I have tried to set 

1 Compare what Dr. Moore says on this matter ; Studies in Dante^ Second 
Series, p. 355. 


forth and from more that other readers of the Concordance may detect, 

I incline strongly to the belief that the chronological order of the four 

works we have been especially considering was V,E,^ Ep, x, A.T,^ M, 

Or at least, assuming this order, as some on other grounds have done, 

we may appeal to stylistic evidence for corroboration. Be that as it may, 

this evidence is enough to refute once and for all the hypothesis that Ep, x 

and A,T, are forgeries; coincidences so numerous and minute could 

have been vouchsafed a forger only by plenary inspiration, proceeding in 

this case from the Father of Lies. As that hypothesis fails, it foUows as 

above stated, that Ep, x and A,T, are genuine works of Dante. The 

remaining letters are too brief to warrant definite conclusions, but I 

would state that I have found no certain evidence against the genuine- 

ness of any one of them. Various coincidences with peculiarities of the 

accepted works will be noted in the lists given above. A similar exam- 

ination of the phraseology of Dante^s Italian prose works might yield 

results of interest. ^ ^^ ^ 

E. K. Rand